Legal Alignment CEO and Vanderbilt adjunct law professor Larry Bridgesmith discusses the state of AI in legal tech today.
Ian Lopez, Legaltech News
While artificial intelligence (AI), in law and beyond, may seem futuristic, the reality is that "it's everywhere," explained Larry Bridgesmith, CEO of Legal Alignment and Vanderbilt University adjunct law professor. Yet while he admits lawyers have "mastered the science off semantics," no one "can define [AI], but everyone is arguing about what it means."
At the 2016 Legal Hackers Summit in Brooklyn, New York, however, Bridgesmith momentarily set aside that question, instead asking technologists and lawyers to "think creatively" about how AI developments in tech "can be a benefit in the world of legal hacking."
According to Bridgesmith, economists look at AI as the "equivalent of a new industrial revolution," while others believe it will lead to a life of leisure "because once the computer brain is equivalent to the human brain, we won't have to work anymore." Yet to those in the know, things aren't quite that exotic.
"Even though it might take a while for AI to reach its potential, I do believe there's a lot going on we can take advantage of if we just open our minds to doing so," he said. "We as lawyers ... much prefer the silver bullet to doing legitimate work." With AI, he added there isn't a silver bullet, but "enormous amounts of ammunition."
So What the Hell is AI in Law?
Digging into the semantic hype that is AI, Bridgesmith described deep machine learning, an ability for a machine to learn without being programmed to learn specific pieces of information. As an example, Bridgesmith pointed to IBM's Watson, to which approved universities, including Vanderbilt, are given access to for hackathons.
In describing deep learning—an area of machine learning that can work through work through multiple levels of abstraction to make sense of data—Bridgesmith pointed to an application called Introspection, which is able to look into an organization's data and find the potential for legal violations before they occur and provide information on how to prevent them from occurring.
Another app is Premonition, which looks into track records of lawyers looking to get hired, providing information on their succession in a particular jurisdiction. Meanwhile, Clause applies "blockchain type technology to create a living contract" with organizational data and is "constantly adjusting" to company relationships. Bridgesmith noted that he's had conversations with leadership at Clause and ROSS, and both companies are open to working with legal hackers.
A more familiar application in law is natural language processing, in which computers build an understanding of natural language. This application is used in e-discovery tools like those from Catalyst, NexLP and Relativity. Bridgesmith also noted companies like i2 Coplink and VIEVU finding patterns in facial recognition.
One company open to working with legal techies is Riverview Law, which has virtual persona assistants, the highest level of which can provide recommendations for alternative ways to approach cases based on available data.
"They're prepared to jump into this pool with us and create an application sandbox for legal hacking," Bridgesmith said.
Additionally, Bridgesmith noted that Neota Logic and HighQ have brought "smart robots" technology, which is frequently used in manufacturing, into law. On July 27, the two companies are set to introduce their "lawyer robot."
This effort, according to HighQ's blog, combines the ability "to pull specific data out of unstructured documents and to move it into a structured format" with replicating "virtually any logical process that can be codified." An example presented in the blog is a law firm's uploading of "a large set of documents," which the technology files then notifies "the appropriate lawyers that the documents are out there waiting for some attention." Additionally, users can be provided with a valuation summary and risk report.
Lastly, Bridgesmith mentioned that video automatic content recognition has the power to find patterns and identities in videos as well as take video testimony and compare it to research on deception.
"All of these things in the etho-sphere in AI ... but the most critically configured area is machine learning applications," Bridgesmith explained, citing information from the Appcessories website on the state of AI. He noted that there are about 260 companies currently operating with this technology. The second-most configured area is natural language processing, while the least is video content recognition.
With $3 billion being spent annually on legal technology, "there's an enormous amount of money and development flowing into this world. We just need to begin to capture it," he said.
No comments:
Post a Comment